Merges
Collection
19 items โข Updated โข 2
How to use knifeayumu/Behemoth-v1.2-Magnum-v4-123B with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="knifeayumu/Behemoth-v1.2-Magnum-v4-123B")
messages = [
{"role": "user", "content": "Who are you?"},
]
pipe(messages) # pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("knifeayumu/Behemoth-v1.2-Magnum-v4-123B")
model = AutoModelForCausalLM.from_pretrained("knifeayumu/Behemoth-v1.2-Magnum-v4-123B", device_map="auto")
messages = [
{"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
tokenize=True,
return_dict=True,
return_tensors="pt",
).to(model.device)
outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))How to use knifeayumu/Behemoth-v1.2-Magnum-v4-123B with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "knifeayumu/Behemoth-v1.2-Magnum-v4-123B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "knifeayumu/Behemoth-v1.2-Magnum-v4-123B",
"messages": [
{
"role": "user",
"content": "What is the capital of France?"
}
]
}'docker model run hf.co/knifeayumu/Behemoth-v1.2-Magnum-v4-123B
How to use knifeayumu/Behemoth-v1.2-Magnum-v4-123B with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "knifeayumu/Behemoth-v1.2-Magnum-v4-123B" \
--host 0.0.0.0 \
--port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "knifeayumu/Behemoth-v1.2-Magnum-v4-123B",
"messages": [
{
"role": "user",
"content": "What is the capital of France?"
}
]
}'docker run --gpus all \
--shm-size 32g \
-p 30000:30000 \
-v ~/.cache/huggingface:/root/.cache/huggingface \
--env "HF_TOKEN=<secret>" \
--ipc=host \
lmsysorg/sglang:latest \
python3 -m sglang.launch_server \
--model-path "knifeayumu/Behemoth-v1.2-Magnum-v4-123B" \
--host 0.0.0.0 \
--port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "knifeayumu/Behemoth-v1.2-Magnum-v4-123B",
"messages": [
{
"role": "user",
"content": "What is the capital of France?"
}
]
}'How to use knifeayumu/Behemoth-v1.2-Magnum-v4-123B with Docker Model Runner:
docker model run hf.co/knifeayumu/Behemoth-v1.2-Magnum-v4-123B
A worthy successor as the v2 didn't set the expectations. SLERPd with less magnum as some people have reported it's being too horny and maybe less coherent.
This is a merge of pre-trained language models created using mergekit.
This model was merged using the SLERP merge method.
The following models were included in the merge:
The following YAML configuration was used to produce this model:
models:
- model: TheDrummer/Behemoth-123B-v1.2
- model: anthracite-org/magnum-v4-123b
merge_method: slerp
base_model: TheDrummer/Behemoth-123B-v1.2
parameters:
t: [0.1, 0.2, 0.4, 0.2, 0.1]
dtype: bfloat16